Harrisburg – Six months after President Donald Trump warned states not to regulate artificial intelligence, they are increasingly doing just that.
Congress has stalled on developing federal regulation on artificial intelligence, while states press ahead, considering how chatbots interact with children, how employers use AI systems, and what steps developers should take to try to avoid an AI catastrophe.
State lawmakers have taken a step back from earlier, more far-reaching attempts to regulate artificial intelligence, which were vetoed or otherwise thwarted by governors who viewed the measures as placing an undue burden on the development of the sector, including efforts to hold developers accountable for biases in artificial intelligence systems.
But they’re coming back with more targeted legislation that often delves into those aspects of life where Americans interact with AI without even realizing it.
Presidential power versus state power
Trump’s move to limit state action on artificial intelligence drew criticism from members of both political parties and civil liberties and consumer rights organizations, who worried that banning state regulation would be a gift to artificial intelligence giants, which are subject to little oversight.
Trump has made artificial intelligence a top priority for national and economic security, saying that allowing states to complicate the regulatory framework for an industry that invests trillions of dollars and drives the economy poses too great a risk in the race with China for AI supremacy.
Trump issued an executive order directing the attorney general to create a task force to challenge state laws that posed “more than minimal” burden, and directing the Commerce Department to develop a list of problematic regulations. Additionally, he threatened to restrict funding from a broadband deployment program and other grant programs to states that contain artificial intelligence laws.
The White House said it would not oppose state laws aimed at preventing fraud and protecting consumers and children.
Meanwhile, the Trump administration released a “national regulatory framework” urging Congress to override state AI laws that do not fit its regulatory vision and pass legislation to protect minors, intellectual property rights, and free speech. In recent days, a new bipartisan draft proposal introduced in the House of Representatives has come under harsh criticism from key figures both Democrats and Republicans.
The White House has shown no signs of following through on its threat to enforce the executive order by going to court against the state’s AI law or withholding funds. In a statement, he stated that the Trump administration is “willing to collaborate with its partners” to implement its regulatory framework.
The states seem to act with hardly any restrictions from Trump
Trump’s executive order does not appear to have deterred states from attempting to regulate the use of artificial intelligence. More bills have been introduced this year than last year, including by Republicans, said Justine Gluck, policy director of the Future of Privacy Forum, a nonprofit organization that advocates for data privacy in technology and whose members come from business, academia and civic groups.
In Illinois, the bill on Democratic Gov. JB Pritzker’s desk draws on elements of laws passed last year in California and New York that require developers of large advanced artificial intelligence models to establish protocols to prevent their systems from causing a catastrophe, such as a biological weapons attack, a power outage or a large-scale cyberattack.
Illinois has introduced a requirement that AI developers must commission an independent auditor to check whether they comply with their own policies. Analysts see it as a step forward in holding AI developers to greater responsibility for their products.
The bill’s sponsor, Democratic state Sen. Mary Edly-Allen, ignored Trump’s threat.
“I don’t know if you know Illinois, but we’re pretty independent,” Edly-Allen told The Associated Press.
The bill received nearly unanimous support, highlighting the willingness of members of Trump’s party to cooperate with Democrats to close the artificial intelligence regulatory gap left by the federal government.
This type of legislation is expected to spread to other states.
The regulation of chatbots, especially for children
More and more states are imposing restrictions on how AI chatbots can interact with people, especially children. Several states, both Republican and Democratic, have passed such laws this year, including Colorado, Connecticut, Idaho, Iowa, Nebraska and Oregon.
In many cases, governments want companies to alert users when they are interacting with an AI rather than a person. Many want the way chatbots interact with minors to be restricted, for parents to have control over their children’s access, and for data provided to chatbots to be kept confidential.
In recent weeks, Connecticut has enacted provisions regarding companion chatbots that maintain an ongoing relationship with people. Under these provisions, chatbots must not be able to interact with minors under 18 years of age, unless they are programmed not to encourage self-destructive behavior and offer parents tools to control minors’ use.
Transparency in AI and decision making
In May, Colorado required companies that use artificial intelligence systems in key areas such as employment, education, housing or banking to inform people when such systems are used to influence a decision that affects them.
It was an attempt to regulate what researchers say is the bias inherent in AI systems that analyze consumer data and make critical decisions, such as who gets hired, who gets a mortgage loan, or who gets medical care. However, the 2024 law, intended to prevent AI’s tendency to discriminate, was watered down under pressure from Democratic Gov. Jared Polis.
In Connecticut, lawmakers have required employers who use artificial intelligence systems in the workplace to inform employees or job applicants that they are interacting with artificial intelligence.
For their part, Connecticut, Washington and Utah have required AI developers to incorporate data into digital content that allows users to determine whether said content – such as photos or videos – has been created or modified by AI.
More laws may be passed this year.
In California, lawmakers are pushing the “Anti-Robot Bosses Act of 2026” to ban employers from relying solely on artificial intelligence when firing or disciplining workers, as well as expanding the state’s regulation of AI chatbots, including a ban on using chatbot responses directed at minors for advertising purposes.
Some Republican-governed states have been reluctant
In Florida, the state House of Representatives refused to greenlight what Republican Gov. Ron DeSantis called his AI “Bill of Rights.” The proposal included provisions to give parents control over their children’s access to companion chatbots, as well as requiring companies that use chatbots to inform consumers when they are interacting with an AI rather than a person.
Florida House Speaker Daniel Perez said Trump had made clear that the federal government should be in charge of regulating artificial intelligence. DeSantis rejected that idea, noting that the federal government is not taking action.
In Utah, progress on a law modeled on regulations in New York and California stalled after the White House sent a one-sentence memo to that state’s lawmakers warning them that it “categorically” opposed the bill.
This story was translated from English to Spanish with an artificial intelligence tool and was reviewed by an editor before publication.